SMART preregistrations Sensible Markup Assists Reviewers Tremendously Tom
SMART pre-registrations Sensible Markup Assists Reviewers Tremendously Tom Hardwicke METRICS Stanford University SIPS 2018, Grand Rapids, Michigan tom. hardwicke@stanford. edu
Traditional research pipeline Data collection Analysis Research paper
Research pipeline with pre-registration Pre-registered protocol Data collection Analysis Research paper
Pre-registration alone is not enough http: //compare-trials. org/ Also see: Chan et al. (2004) JAMA Mathieu et al. (2009) JAMA Walker et al. (2014) JRSMO Zarin et al. (2011) NEJM Chan et al. (2017) JAMA
Reviewers not checking pre-registered protocols Survey: Of 672 respondees who had reviewed clinical trials in the last 2 years, only 232 said that they had examined the registered protocol (34. 3%).
Peer-review of protocols and papers is like an unpleasant game of spot the difference Pre-registered protocol Research paper Peer-review
Navigating the modern office environment Beesley, P. , Halpert, J. Schrute, D. , & Scott, M. We recruited a sample of 61 participants as specified by our pre-registered stopping rule
SMART pre-registered protocol: Navigating the modern office environment Beesley, P. , Halpert, J. Schrute, D. , & Scott, M. P-1: We hypothesized that participants in the beetroot condition would have better mood scores than participants in the no-beetroot control condition. P-2: The dependent variable was mood scores as measured on a 5 -point likert scale ranging from “feeling a bit gloomy” through to “smashing! Thanks” ……. . P-6: A power analysis was carried out to determine that a sample size of 61 participants was necessary to detect medium effects f =. 25, as outlined by Cohen, 1988), with power of. 9.
e p a p e In th r We recruited a sample of 61 participants[P-6] e h t m stered o r F gi e r pre l o c o t pro P-6: A power analysis was carried out to determine that a sample size of 61 participants was necessary to detect medium effects f =. 25, as outlined by Cohen, 1988), with power of. 9.
SMART pre-registrations Sensible Markup Assists Reviewers Tremendously Tom Hardwicke tom. hardwicke@stanford. edu
- Slides: 11